Research On Evolutionary Rule Design And Evaluation Technology Oriented To Software Evolution Scenarios

Software evolution is an update behavior that runs through the entire life cycle of software.These behaviors will inevitably introduce changes and bring new risks.The existence of these potential risks has led to the analysis and evaluation of evolution.In actual engineering,requirements are the factors that drive evolution.Different requirements generate different evolution scenarios,and the goals of software evolution and changes within the software will also be different under different evolution scenarios.However,most of the existing evolution assessment studies do not focus on evolution scenarios,and their main focus is on which quality attributes may be affected by the evolution process.The main purpose is to ensure the quality of software,which leads to evolution assessment results in different scenarios.The degree of discrimination is not high and cannot provide evaluation results from multiple angles.In order to solve the above problems,this thesis studies and designs evolution rules and rulebased evolution evaluation for software evolution scenarios.This thesis analyzes the evolution goals and evolution characteristics of software in different evolution scenarios driven by different requirements by studying the requirements that drive software evolution,and designs targeted software evolution scenarios based on different evolution scenarios,evolution goals and evolution characteristics.According to the evolution rules,the measurable and testable data that can reflect the evolution characteristics of the software are selected as the evaluation objects,and the corresponding measurement and evaluation formulas are designed according to the evolution rules to effectively quantify the evaluation results.On this basis,this thesis designs a rule-based evolution evaluation process,including obtaining the data required for evolution evaluation,reading the original version and evolution version of the software,and performing evolution evaluation.Finally,this thesis designs and implements the corresponding software evolution evaluation tool SEAT,which implements the evolution evaluation rules and evolution evaluation technology designed in this thesis,can evaluate the evolution of the software to be evaluated,and give the evaluation results.This thesis selects sixteen versions of six open source projects under three evolution scenarios:perfect,adaptive and preventive to design experiments to verify the information extraction function of SEAT tool and the validity of the evolution rules designed in this thesis,and to verify the validity of the evolution rules designed in this thesis.Analyze the characteristics of software evolution in different scenarios.The experimental results show that the SEAT tool has high accuracy for the identification of classes,interfaces,methods,etc.In addition,the evaluation results of the evolution rules designed in this thesis are in line with the actual evolution of the software,and the evaluation results can better reflect the software Evolution effects and internal changes in different scenarios,so as to help developers better understand,control and adjust evolution information and evolution trends related to evolution scenarios.